Prediction Model for Audiological Outcomes in Patients With GJB2 Mutations

Recessive mutations in GJB2 are the most common genetic cause of sensorineural hearing impairment (SNHI) in humans. SNHI related to GJB2 mutations demonstrates a wide variation in audiological features, and there has been no reliable prediction model for hearing outcomes until now. The objectives of this study were to clarify the predominant factors determining hearing outcome and to establish a predictive model for SNHI in patients with GJB2 mutations.

A total of 434 patients confirmed to have biallelic GJB2 mutations were enrolled and divided into three groups according to their GJB2 genotypes. Audiological data, including hearing levels and audiogram configurations, were compared between patients with different genotypes. Univariate and multivariate generalized estimating equation (GEE) analyses were performed to analyze longitudinal data of patients with multiple audiological records.

Of the 434 patients, 346 (79.7%) were homozygous for the GJB2 p.V37I mutation, 55 (12.7%) were compound heterozygous for p.V37I and another GJB2 mutation, and 33 (7.6%) had biallelic GJB2 mutations other than p.V37I. There was a significant difference in hearing level and the distribution of audiogram configurations between the three groups. Multivariate GEE analyses on 707 audiological records of 227 patients revealed that the baseline hearing level and the duration of follow-up were the predominant predictors of hearing outcome, and that hearing levels in patients with GJB2 mutations could be estimated based on these two parameters: (Predicted Hearing Level [dBHL]) = 3.78 + 0.96 × (Baseline Hearing Level [dBHL]) + 0.55 × (Duration of Follow-Up [y]).

The baseline hearing level and the duration of follow-up are the main prognostic factors for outcome of GJB2-related SNHI. These findings may have important clinical implications in guiding follow-up protocols and designing treatment plans in patients with GJB2 mutations.